检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Wang, Zhenglin Wu, Jialong Lai, Yilong Zhang, Congzhi Zhou, Deyu School of Computer Science and Engineering Key Laboratory of Computer Network and Information Integration Ministry of Education Southeast University China

Large Language Models (LLMs) demonstrate remarkable emergent abilities across various tasks, yet fall short of complex reasoning and planning tasks. The tree-search-based reasoning methods address this by encouraging the exploration of intermediate steps, surpassing the capabilities of chain-of-thought prompting. However, significant inference latency is introduced due to the systematic exploration and evaluation of multiple thought paths. This paper introduces SEED, a novel and efficient inference framework to improve both runtime speed and GPU memory management concurrently. Based on a scheduled speculative execution, SEED efficiently handles multiple iterations for thought generation and state evaluation, leveraging a rounds-scheduled strategy to manage draft model dispatching. Extensive experimental evaluations on three reasoning datasets demonstrate the superior speedup performance of SEED1 Copyright © 2024, The Authors. All rights reserved.

关键词： Memory management

来源：评论

学校读者我要写书评

暂无评论

Causal Walk: Debiasing Multi-Hop Fact Verification with Front-Door Adjustment

arXiv

引用

arXiv 2024年

作者： Zhang, Congzhi Zhang, Linhai Zhou, Deyu School of Computer Science and Engineering Key Laboratory of Computer Network and Information Integration Ministry of Education Southeast University China

Multi-hop fact verification aims to detect the veracity of the given claim by integrating and reasoning over multiple pieces of evidence. Conventional multi-hop fact verification models are prone to rely on spurious correlations from the annotation artifacts, leading to an obvious performance decline on unbiased datasets. Among the various debiasing works, the causal inference-based methods become popular by performing theoretically guaranteed debiasing such as casual intervention or counterfactual reasoning. However, existing causal inference-based debiasing methods, which mainly formulate fact verification as a single-hop reasoning task to tackle shallow bias patterns, cannot deal with the complicated bias patterns hidden in multiple hops of evidence. To address the challenge, we propose Causal Walk, a novel method for debiasing multi-hop fact verification from a causal perspective with front-door adjustment. Specifically, in the structural causal model, the reasoning path between the treatment (the input claim-evidence graph) and the outcome (the veracity label) is introduced as the mediator to block the confounder. With the front-door adjustment, the causal effect between the treatment and the outcome is decomposed into the causal effect between the treatment and the mediator, which is estimated by applying the idea of random walk, and the causal effect between the mediator and the outcome, which is estimated with normalized weighted geometric mean approximation. To investigate the effectiveness of the proposed method, an adversarial multi-hop fact verification dataset and a symmetric multi-hop fact verification dataset are proposed with the help of the large language model. Experimental results show that Causal Walk outperforms some previous debiasing methods on both existing datasets and the newly constructed datasets. Code and data will be released at https://***/zcccccz/CausalWalk. Copyright © 2024, The Authors. All rights reserved.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

Inaccurate Label Distribution Learning with Dependency Noise

arXiv

引用

arXiv 2024年

作者： Kou, Zhiqiang Wang, Jing Jia, Yuheng Geng, Xin MOE Key Laboratory of Computer Network and Information Integration School of Computer Science and Engineering Southeast University Nanjing China

In this paper, we introduce the Dependent Noise-based Inaccurate Label Distribution Learning (DN-ILDL) framework to tackle the challenges posed by noise in label distribution learning, which arise from dependencies on instances and labels. We start by modeling the inaccurate label distribution matrix as a combination of the true label distribution and a noise matrix influenced by specific instances and labels. To address this, we develop a linear mapping from instances to their true label distributions, incorporating label correlations, and decompose the noise matrix using feature and label representations, applying group sparsity constraints to accurately capture the noise. Furthermore, we employ graph regularization to align the topological structures of the input and output spaces, ensuring accurate reconstruction of the true label distribution matrix. Utilizing the Alternating Direction Method of Multipliers (ADMM) for efficient optimization, we validate our method's capability to recover true labels accurately and establish a generalization error bound. Extensive experiments demonstrate that DN-ILDL effectively addresses the ILDL problem and outperforms existing LDL methods. © 2024, CC BY.

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

Compositional metric learning for multi-label classification

引用

Frontiers of computer Science 2021年第5期15卷 1-12页

作者： Yan-Ping SUN Min-Ling ZHANG School of Computer Science and Engineering Southeast UniversityNanjing 210096China Key Laboratory of Computer Network and Information Integration(Southeast University) Ministry of EducationChina Collaborative Innovation Center for Wireless Communications Technology Nanjing 211100China

Multi-label classification aims to assign a set of proper labels for each instance,where distance metric learning can help improve the generalization ability of instance-based multi-label classification *** multi-label metric learning techniques work by utilizing pairwise constraints to enforce that examples with similar label assignments should have close distance in the embedded feature *** this paper,a novel distance metric learning approach for multi-label classification is proposed by modeling structural interactions between instance space and label *** one hand,compositional distance metric is employed which adopts the representation of a weighted sum of rank-1 PSD matrices based on com-ponent *** the other hand,compositional weights are optimized by exploiting triplet similarity constraints derived from both instance and label *** to the compositional nature of employed distance metric,the resulting problem admits quadratic programming formulation with linear optimization complexity *** number of training *** also derive the generalization bound for the proposed approach based on algorithmic robustness analysis of the compositional *** experiments on sixteen benchmark data sets clearly validate the usefulness of compositional metric in yielding effective distance metric for multi-label classification.

关键词： machine learning multi-label learning metric learning compositional metric positive semidefinite matrix decomposition

来源：评论

学校读者我要写书评

暂无评论

ADACQR: Enhancing Query Reformulation for Conversational Search via Sparse and Dense Retrieval Alignment

arXiv

引用

arXiv 2024年

作者： Lai, Yilong Wu, Jialong Zhang, Congzhi Sun, Haowen Zhou, Deyu School of Computer Science and Engineering Key Laboratory of Computer Network and Information Integration Ministry of Education Southeast University China

Conversational Query Reformulation (CQR) has significantly advanced in addressing the challenges of conversational search, particularly those stemming from the latent user intent and the need for historical context. Recent works aimed to boost the performance of CQR through alignment. However, they are designed for one specific retrieval system, which potentially results in sub-optimal generalization. To overcome this limitation, we present a novel framework ADACQR. By aligning reformulation models with both term-based and semantic-based retrieval systems, ADACQR enhances the generalizability of information-seeking queries among diverse retrieval environments through a two-stage training strategy. Moreover, two effective approaches are proposed to obtain superior labels and diverse input candidates, boosting the efficiency and robustness of the framework. Experimental results on the TopiOCQA, QReCC and TREC CAsT datasets demonstrate that ADACQR outperforms the existing methods in a more efficient framework, offering both quantitative and qualitative improvements in conversational query reformulation.1 Copyright © 2024, The Authors. All rights reserved.

关键词： Structured Query Language

来源：评论

学校读者我要写书评

暂无评论

Deep Iterative Reconstruction network Based on Residual Constraint for Low-Dose CT Imaging 8

Deep Iterative Reconstruction Network Based on Residual Cons...

引用

8th International Conference on Systems and Informatics, ICSAI 2022

作者： Liu, Jin Kang, Yanqin Liu, Tao Zhang, TingYu Zhang, Yikun Anhui Polytechnic University School of Computer and Information Wuhu China Southeast University Ministry of Education Key Laboratory of Computer Network and Information Integration Nanjing China Southeast University School of Computer Science and Engineering Nanjing China

ISBN: (纸本)9781665474481

clinical low X-ray dose computed tomography (LDCT) scanner often induce high intensity strip artifact and spot nosie, compromising diagnoses and intervention plans. Recently, sparsely constrained and network learning-based frameworks have been shown to be efficient in mitigating such issue. In this work, we propose a deep iterative reconstruction network (DIRNet) model with a residual constraint to synergize the advantages of feature learning and image reconstruction to address the LDCT imaging problem. DIR-Net compose by few iteration units, and all iteration units include three different network modules: projection restoration, residual constraint and image update block. DIR-Net is a promising approach for building an end-to-reconstruction mapping strategy and directly obtaining high-quality CT images. Furthermore, LISTA is used to conFigure the network, and the whole network architecture yields improved interpretability. Qualitative and quantitative analysis in test data shown the promising imaging effects of DIR-Net in quantum noise reduction, block artifact removal and tissue detail texture mantian. © 2022 IEEE.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Binary decomposition: a problem transformation perspective for open-set semi-supervised learning 24

Binary decomposition: a problem transformation perspective f...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Jun-Yi Hang Min-Ling Zhang School of Computer Science and Engineering Southeast University Nanjing China and Key Laboratory of Computer Network and Information Integration Southeast University Ministry of Education China

Semi-supervised learning (SSL) is a classical machine learning paradigm dealing with labeled and unlabeled data. However, it often suffers performance degradation in real-world open-set scenarios, where unlabeled data contains outliers from novel categories that do not appear in labeled data. Existing studies commonly tackle this challenging open-set SSL problem with detect-and-filter strategy, which attempts to purify unlabeled data by detecting and filtering outliers. In this paper, we propose a novel binary decomposition strategy, which refrains from error-prone procedure of outlier detection by directly transforming the original open-set SSL problem into a number of standard binary SSL problems. Accordingly, a concise yet effective approach named BDMatch is presented. BDMatch confronts two attendant issues brought by binary decomposition, i.e. class-imbalance and representation-compromise, with adaptive logit adjustment and label-specific feature learning respectively. Comprehensive experiments on diversified benchmarks clearly validate the superiority of BDMatch as well as the effectiveness of our binary decomposition strategy.

关键词：

来源：评论

学校读者我要写书评

暂无评论

PROMIPL: A Probabilistic Generative Model for Multi-Instance Partial-Label Learning

PROMIPL: A Probabilistic Generative Model for Multi-Instance...

引用

IEEE International Conference on Data Mining (ICDM)

作者： Yin-Fang Yang Wei Tang Min-Ling Zhang School of Computer Science and Engineering Southeast University Nanjing China Key Laboratory of Computer Network Information Integration (Southeast University) Ministry of Education China

ISBN: (数字)9798331506681

ISBN: (纸本)9798331506698

Multi-instance partial-label learning (MIPL) tackles scenarios where each training sample is represented as a multiinstance bag associated with a candidate label set. This set contains one true label and several false positives. Existing MIPL algorithms have predominantly focused on mapping multiinstance bags to candidate label sets for disambiguation. However, these algorithms may not be adequately generalizable in intricate real-world situations due to their reliance on heuristic methods for identifying true labels. In this paper, we propose PROMIPL, i.e., a PRObabilistic generative model for Multi-instance partiallabel learning, to address these challenges. PROMIPL is the first attempt to explore the probabilistic generative model to infer latent ground-truth labeling information from the data generation process in multi-instance partial-label learning. Besides, the discovered underlying structures also provide improved explanations of the classification predictions. To circumvent the computationally intensive process of training the generative model, we formulate a unified variational lower bound within the stochastic gradient variational Bayesian framework for the model parameters. Experimental results from benchmark and realworld datasets show that our proposed PROMIPL is competitive or superior to the state-of-the-art methods.

关键词： Training Lower bound Computational modeling Heuristic algorithms Stochastic processes Probabilistic logic Data models Bayes methods Labeling Data mining

来源：评论

学校读者我要写书评

暂无评论

TPMARL-MSS Thought-Perception Multi-Agent Reinforcement Learning Meeting Scheduling System 25

TPMARL-MSS Thought-Perception Multi-Agent Reinforcement Lear...

引用

25th IEEE International Conferences on High Performance Computing and Communications, 9th International Conference on Data Science and Systems, 21st IEEE International Conference on Smart City and 9th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Applications, HPCC/DSS/SmartCity/DependSys 2023

作者： Zhang, Jinsong Mao, Xiaoxuan Li, Lu Wu, Hanqian Joint Graduate School Southeast University Nanjing China College of Software Engineering Southeast University Nanjing China School of Cyber Science and Engineering Southeast University Nanjing China School of Cyber Science and Engineering Southeast University Key Laboratory of Computer Network and Information Integration of Ministry of Education Nanjing China

ISBN: (纸本)9798350330014

In today's evolving society, the increasing complexity and frequency of meetings necessitate advanced scheduling systems. Traditional methods are constrained by rigid prede-fined strategies, lack intelligent negotiation mechanisms, and often compromise user privacy. Addressing these challenges, we introduce the Thought-Perception Multi-Agent Reinforcement Learning Meeting Scheduling System (TPMARL-MSS). Unlike conventional systems, TPMARL-MSS autonomously learns and refines its strategies through continuous feedback. It features automated negotiation and adaptive decision-making, offering a more nuanced scheduling approach. Importantly, our Thought-Perception module protect privacy, allowing the agent to deduce preferences from participant behavior without revealing personal data. Evaluations on the real-world dataset shows that TPMARL-MSS surpasses traditional methods in efficiency and schedule quality, highlighting its practical applicability. © 2023 IEEE.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Early Prediction of Heart Disease via LSTM-XGBoost 23

Early Prediction of Heart Disease via LSTM-XGBoost

引用

9th International Conference on Computing and Artificial Intelligence, ICCAI 2023

作者： Zang, Xiaodong Du, Jin Song, Yuansheng School of Cyber Science and Engineering Qufu Normal University Key Laboratory of Computer Network and Information Integration Ministry of Education Southeast University Nanjing211189 China School of Cyber Science and Engineering Qufu Normal University China

ISBN: (纸本)9781450399029

With the development of information and technology, especially with the boom in big data, healthcare support systems are becoming much better. However, an early diagnosis is not an easy task because it is hard to find out the hidden patterns in the disease. The symptoms of heart failure are usually non-specific. Therefore, this study aims to devise an early prediction model for incident heart disease based on machine learning algorithms. In this study, we compared several powerful models, including support vector machines (SVM), k-nearest neighbor (KNN), decision tree, extreme gradient boosting (XGBoost), random forest, gradient boosting decision tree (GBDT), categorical boosting (CatBoost), light gradient boosting machine (LightGBM), and choose two best models, namely, XGBoost and LSTM as our prediction model. The XGBoost algorithm was used to extract a subset from the data of patients, which will be fed into the LSTM-XGBoost model for training. An experiment comparison is conducted based on the data from Kaggle. Experimental results demonstrate that the LSTM-XGBoost model can operate effectively with a prediction accuracy achieving 0.9896 and a loss rate achieving 0.0105, which is much better than XGBoost or LSTM, respectively. © 2023 Copyright held by the owner/author(s).

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：